Deposing Dictators

[Arthur_Siegel at rsmi.com]

Terry J. Reedy writes -

>The division change (and some arguments pro and con) was formulated at
>least by November 98 (see 'Deja/vu').  Did VPython exist then?  (I
>have no idea.)

I believe that that discussion was in fact pre VPython, but post Alice.

The airtight case put forth is that the significance of VPython was in 
confirming
Alice's reports.

And the truth of the matter is that, stand-alone at least, the Alice concerns 
on the
div operator are credible and internally consistent.  It is geared to folk 
working
totally outside the realm of math, programming, science etc. Its whole ethic is

to 
insulate its users from complexity, at all and  any cost. 

Fork sounds right for them, IMO. 

The VPython reports particularly concern me because we are now saying that 
science
students working in the strictly numeric realm, with the Numeric tool, in which

is 
embedded a slew of rules and behaviors regarding the integer numeric type - 
has similar needs and experience as the Alice audience. Ain't so.  Anybody 
working
with VPython *needs* to understand the significance of numeric types.  And once

again,
the old div operator behavior was a great signpost and the least of their 
problems  -
particularly if they are going into it without clear instruction on what to 
expect.

It *will* be different in the unified numerical model days. But that ain't
next
semseter, for sure.

ART